Tianqi Chen comments

Results 637 comments of


                                            Tianqi Chen

[Tracking Issue] How to avoid TVM increased tensor dimension?

for now we don't have such setting unfortunately

[Question] Does Android MLCChat has previous context as input?

Thanks for reporting, we are working to address this as https://github.com/mlc-ai/mlc-llm/issues/2740

[Codegen, CUDA] Enable emitting SyncWarp

@tvm-bot rerun

[Codegen, CUDA] Enable emitting SyncWarp

@vinx13 do you mind doing a rebase? seems the rerun logic may not work too well for a merge commit

[Codegen, CUDA] Enable emitting SyncWarp

hmm, seems indeed a timeout, so maybe there is someissue in allreduce gen (the ci gpu is T4), not sure if it is related to gpu types, leave it here...

[Question] How to use cpp in project

If you are interested in using the c++ api, perhaps the easiest way is to leverage the JSON FFI. You can try to dig into the Swift binding as an...

[CPP_CLI] MLC Cli App over JSONEngine interface

@mengshyu can you also help to review this pR

[RELAX][PASS] Annotate Custom Scope layout pass for Adreno GPU

@Hzfengsy do you mind take a look given it touches FuseOps/TIR

[RELAX][PASS] Annotate Custom Scope layout pass for Adreno GPU

also cc @yongwww for memory scope related changes

[RELAX][PASS] Annotate Custom Scope layout pass for Adreno GPU

Thanks @srkreddy1238 for updates. I take a closer look and now understands the motivation behind `add_attributes`. This is mainly to handle the case of conv2d operators where texture can be...